Use LEFT and RIGHT arrow keys to navigate between flashcards;
Use UP and DOWN arrow keys to flip the card;
H to show hint;
A reads text to speech;
81 Cards in this Set
- Front
- Back
Quasi experiment
|
type of research design where a comparison is made, as in the experiment, but no random assignment of participants to groups occurs
|
|
random assignment
|
participants(P) are randomly assigned to levels of the independent variable(IV) in an experiment to control for the individual differences as an extraneous variable.
|
|
Quasi-independent variables(V)
|
variables that cannot be manipulated. therefore not true variables
|
|
Pretest-posttest Design:
|
compares treatments before and after.
- can include a control group that does not have the treatment - Quasi experimental due to non-random assignment (Non equivalent groups) |
|
If you do find improvement over the years what is the factor of this change? (Pretest-posttest design)
- What can it be due to? The treatment or some extraneous factor - To prevent this we throw in a control group. |
throw in a control group
|
|
quasi experiments have
Sources of Bias: |
1. testing effects
2. history effects 3. regression toward the mean 4. maturation: natural changes that occur to the participant during the course of a study that can result in bias |
|
Solomon Four-group design
|
pretest posttest design with two sets of nonequivalent
• A second group is done without a pretest • It is used when you are concerned about testing effects |
|
Time series design(TSD)
|
Measure people several time through a long period of time
In the middle of it you have treatment • Patterns of score before and after treatment |
|
TSD 1. Interrupted design:
|
naturally occurring treatment
|
|
TSD 2.Non-interrupted design:
|
treatment implemented by researcher
|
|
history effects
|
events that occur during the course of the study to all or individual P(s) that can result in bias
|
|
Nonequivalent groups
|
groups compared in a experiment where P(s) are not randomly assigned
|
|
regression toward the mean
|
can occur when P(s) score higher or lower than there personal average. the next time they are tested they are inclined to score their average, making scores unreliable
|
|
Testing effects
|
occur when P(s) are tested more than once in a study with early testing effects affecting later testing
|
|
Developmental designs that treat the factor of age differently
|
1. Longitudinal design (LD)
2.Cross-sectional design (CSD) 3.Cohort sequential design |
|
LD
|
treats age as within subjects variable
P(s) are tested at different ages of their lives a dev. design(d) where a single sample of P(s) is followed over time and tested at diff. ages |
|
within subjects variable
|
each P(s) experiences all levels of the variable
|
|
problems with LD?
|
-testing effects: e.g. want to see how ppl political views change as they are gong through college, so we test them when they come in freshman year. We might answer later in the future in a way that seem consistent with our questions asked.
-Mortality/attrition: when ppl drop out of the experiment without finishing the entire experiment. |
|
CSD
|
treat age as between subjects variable.
this design compare diff. agr groups of P(s) where each P contributes data for only one age group. |
|
Between Subjects Variable
|
each P experiences only one level of the IV
|
|
CSD
|
book definition: a dev. D where multiple samples of P(s) of diff ages are tested once
|
|
CSD good stuff
|
test P(s) only once eliminating attrition
also, it doesnt take a long period of time to complete it. |
|
CSD Biases:
|
confound that may occur in CSD due to diff. experiences that diff, generations have
- History effects in a cross-sectional decline could be influenced by the type of education they have received. NO individual difference in LD |
|
Cohort sequential designs
|
- Measuring age within subjects variable and also a between subject design
- First start with a CSD of three different ages then take these same groups and follow them with a LD. Essentially what one is doing is controlling for both biases. - Hopefully what you find is that when testing three different groups over time will produce similar effects |
|
Small n design
|
Test one or few individuals in experiment or quasi experiment to better understand the behavior .
|
|
baseline measurement
|
a measurement of behavior without a treatment used as a comparison
|
|
discrete trials design
|
a small n design that involves a large number of trials completed by one or few individuals and conducted to describe basic behaviors
|
|
baseline designs
|
s small n design that involves baseline measurements of behaviors as compared with measures of behaviors during the implementation of a treatment
|
|
ABA/reversal design
|
baseline bahavior is measured, followed by implementation of treatment, follwed by another baseline measure after the treatmemnt has stopped
|
|
DISCUSSION SECTION APA
|
- In your introduction you did a literature review of 5 articles
- The results section just tells the statistical results and the discussion talks about the results. It mirrors what we found in introduction along with other studies. - Give brief findings of study don’t put any statistical data. E.g. might start with - In the present study (the one that we are presenting) and give the main findings…it was shown that males do better then females in a low stress. Probably one sentence or two. - At the end of intro you gave your hypothesis, now compare results with the hypothesis. In most cases your results wont confirm the hypothesis (H). Terms used: it supports H or it does not support H. can never say we proved something. - Compare results with previous literature. Our results differ from…that found males did better in high anxiety condition. - If results differ from others discuss why our results were different from others. - Hardest part: discuss what theoretical impact results have on the topic st |
|
table
|
- in a presentation it is good to make a graph for the class.
- Don’t do both in a presentation. Only choose to do the data once. - Comes right after the references. - You a single page for each table - Cannot copy and paste table from SPSS, you must type and make it look like APA tables (129-150) - The graphs on SPSS usually start at 60 which should not be done that way. One should always start with zero. Do a graph on Microsoft excel. |
|
Appendix
|
- very end of the paper
- questionnaire - copy of pictures - script - how do we refer to it, participant filled out inventory (see appendix label them with A,B,C,D) should be on the materials section. |
|
abstract
|
- should be the last thing that we should write because it is a summary of the entire paper
- Less then one page, less then 250 words. - 1st state research question - Nature of participant sample brief description of methods used and designed used (one sentence/ combined together). - Statement if main findings with no statistics. - statements of conclusions drawn. - implications or applications of your findings. - title is abstract and centered - page two of manuscript -Do not indent any of the lines one long paragraph |
|
Sampling error
|
the difference between the observation in a population and in the sample that represents that population in a study
|
|
Central tendency
|
representation of a typical score in a distribution (e. mean, median, and mode)
|
|
mean:
median: mode: |
me: the calculated average of the scores in a distribution.
med: the middle score in the distribution such that half of the sores are above and the other half are below thta value mo: the most common score in a distribution |
|
outliers
|
extreme scores affect the mean
|
|
reaction time
|
measurement of the length of tie to complete a task
|
|
variability
|
how much scores inthe distribution differ from each other across the response scale.
e.g. on a 1-5 scale if the Ps only used 2-4 then there is a low variability meaning that the score are mostly the same |
|
range
|
difference between the highest scpre and the lowest score
|
|
SD
|
the average difference between the scores and the mean of a distribution
|
|
n-1
|
degrees of freedom: the number of scores that can vary in the calculation of a statistic
|
|
variance
|
the SD squared
|
|
predictor variable
|
IN a correlational studies a variable is used to predict the score in another varibale
|
|
outcome variable
|
the DV in a correlational study that is being predicted by the predictor
|
|
Inferential statistics
|
a set of statistical procedures used by researchers to test hypotheses about populations
|
|
two tailed hypothesis
|
both directions of an effect or relationship are considered in the alternative hypothesis of the test
|
|
one-tailed hypothesis
|
only one direction of an effect or relationship is predicted in the alternative hypothesis of the test
|
|
distribution of sample means
|
the distribution of all possible sample means for all possible samples from a population.
|
|
significance testing
alpha level |
the probabilty level used by researchers to indicate the cuttoff probabilty level(highest value) that will allow them to reject the null hypothesis
|
|
ST
p value |
probability value with inferentil statistics that indicates the likelihood of obtaining the data in a study when null hypo is true
|
|
type 1 error
|
reject the null but it is actually correct
|
|
type 2 error
|
fail to reject the null hypo. when it is actually false
|
|
by keeping the typw two error rate low you are increasing the
|
Power of your significance test to detect an effect or relationship that actually exist
|
|
true independent variable
|
a V in an experi. that is manipulated by the researcher such that the levels of the V changes across or within subjects in the Experiment.
|
|
good manipulation of IV increases the...good test of causal hypothesis
|
internal validity
|
|
3rd variable problem
|
the presence of extraneous factors in a study that affect the DV can decrease the internal validity of the study
|
|
confounding variables
|
an extraneous factor present in thestudy that may affect results
|
|
Factorial design
|
an experiment or quasi experiment that includes more tha one independent variable.
|
|
main effects
|
test of the differences between all means for each level of an independent variable in an ANOVA
|
|
interaction effects
|
test the effects of one independent variable at each level of another independent variable in an ANOVA
|
|
simple effects test
|
statistical tests conducted to charcaterize an interraction effect when one is found in an ANOVA
|
|
null hypo
|
the hypo that an effect or relationship does not exist (or exist in the direction opposite of the alternative hypo)
|
|
critical region
|
the most extreme portion of a distribution of statistic values for the null hypo, determined by alpha level 5%
|
|
alternative hypo
|
the hypo that a relationship or an effect exist in the population
|
|
significant
|
the p value is less than or equal to alpha in an inferential test, null hypo can be rejected
|
|
if variability is high...
|
it is likely to contain sampling error
|
|
t test
|
significance test used to compare means
|
|
one sample t test
|
used when the population mean without the treatment is known and is compared with a single sample
|
|
independent samples t test
|
when 2 samples w diff. individuals are compared
|
|
repeated measures samples t test
|
when two related sample or two related scores from same individual are compared
|
|
ANOVA
|
when more than 2 samples or sets of scores from the same individual are compared
|
|
pearson r test
|
when a relationship between two sets of scores is being tested
|
|
regression
|
when you want to predict an individuals score on one variable from the score on a second, related variable
|
|
multivalent variable
|
one variable many levels
|
|
factorial design
|
an experi. or quasi experi. that includes more than one IV
|
|
ANOVA
|
analysis of variance test used for designs with 3 or more sample means
|
|
main effect
|
test the difference betwen all means for each level of the independent variable in an ANOVA
|
|
post hoc test
|
additonal significance test conducted to determine which means are significantly diff. for a main effect
|
|
interaction effect
|
test the effect of one IV at each level of another IV
|
|
linear regression
|
statitical technique that determines the best fit line to a set of data to predict the score on one variable from the score of another
|